targets & mechanisMS
Second wind for T cells By Lev Osherovich, Senior Writer In the battle against HIV, HBV and HCV, all of which establish persistent infections that eventually wear down the antiviral cytotoxic T cell response, the odds for the immune system are poor. Three studies now show that blocking a bevy of T cell–inhibitory receptors can reverse this effect. The findings provide new validation for the approach taken by Medarex Inc., which has four products targeting such receptors in its cancer and HCV pipelines. Two studies, published in Nature Immunology and the Proceedings of the National Academy of Sciences, used a murine lymphocytic choriomeningitis virus (LCMV) model of chronic viral infection to identify receptors that suppress cytotoxic T lymphocyte (CTL) response. 1,2 A third study in Nature showed the efficacy of the strategy in a macaque model of HIV.3
Inhibitory receptors, which include cytotoxic T lymphocyte–associated protein 4 (CTLA4; CD152) and the related protein programmed cell death 1 (PDCD1; CD279; PD-1), ordinarily function to dampen excessive T cell activation that can lead to autoimmune disease. However, chronic viral infection prompts T cells to upregulate CTLA4 and PD-1, thus undermining CTL activity and gradually driving T cells into a quiescent state termed exhaustion. Exhausted CTLs can no longer respond to virally infected host cells, leaving viruses free to replicate. Losing inhibition The Nature Immunology study identified surface receptors that work in parallel with PD-1 and CTLA4 to inhibit CTLs. These receptors could potentially be antagonized to increase the response to viral infection. The research team had suspected that PD-1 and CTLA4 were not the only receptors that can turn down CTL responses. E. John Wherry, assistant professor of immunology at The Wistar Institute and lead author on the paper, told SciBX that previous efforts to treat viral infections by blocking either PD-1 or CTLA4 “improved the response, but not entirely.” To identify receptors with similar roles to PD-1 and CTLA4, the researchers used bioinformatics to create a list of candidate genes that are upregulated along with PD-1 in chronic viral infection. “We used gene expression profiling to identify seven receptors and found that these receptors work together to fine tune” negative regula-
Table 1. Targeting inhibitory receptors to boost immunity. Blackburn et al. and Brooks et al. used a murine lymphocytic choriomeningitis virus (LCMV) model to identify receptors on cytotoxic T lymphocytes (CTLs) that could be targeted to increase the immune response to chronic viral infections such as HIV, HBV and HCV. This table summarizes the human homologs of potential targets described in the two studies. Unless otherwise noted, compounds mentioned below bind to the target. Target name
Description
Related compounds
CD160
Glycoprotein on CTL and peripheral blood natural killer (NK) cells
MAT302 antibody in preclinical development for solid tumors and wet age-related macular degeneration (AMD) from MAT Biopharma S.A.
CD244; NK cell receptor 2B4
Glycoprotein expressed on the surface of CTLs and NK cells
Not applicable
CTL-associated protein 4 (CTLA4; CD152)
Transmembrane receptor that sends an inhibitory signal in activated T cells
Orencia abatacept, a CTLA4-Ig fusion protein that binds to B7-1 (CD80) and B7-2 (CD86), on the market for rheumatoid arthritis from Bristol-Myers Squibb Co. (NYSE:BMY); ipilimumab (MDX010) in Phase III for melanoma from Medarex Inc. (NASDAQ: MEDX)/Bristol-Myers; tremelimumab (CP-675206) in Phase II for various cancers from Pfizer Inc. (NYSE:PFE)
IL-10 receptor
Receptor for IL-10; expressed on a variety of lymphocytes
IL-10 antibody from Schering-Plough Corp. (NYSE:SGP) used in the PNAS study by Brooks et al.
Leukocyte immunoglobulin-like receptor, subfamily B (with TM and ITIM domains), member 3
Transmembrane protein expressed on immune cells that binds major histocompatibility complex, class I molecules and sends inhibitory signals
Not applicable
Lymphocyte-activation gene 3 (LAG3; CD223)
Transmembrane receptor expressed on T cells; resembles CD4
Target of a preclinical program at Medarex; ImmuTune IMP701 for cancer and ImmuTune IMP731 for autoimmune diseases and transplantation in preclinical development from Immutep S.A.
Programmed cell death 1 (PDCD1; PD-1; CD279)
Transmembrane receptor that inhibits B cell and T cell activity
MDX-1106 antibody in Phase I for solid tumors and HCV infection from Medarex/Ono Pharmaceuticals Co. Ltd. (Tokyo:4528; Osaka:4528)
CD274 (PD-L1)
Ligand for PD-1 expressed on antigen-presenting cells and many tumor cells
MDX-1105 antibody in Phase I for solid tumors from Medarex
SciBX: Science–Business eXchange
Copyright © 2008 Nature Publishing Group
targets & mechanisMS tion of CTLs, said Wherry. “Almost all of the new targets we identified Scripps, and David Brooks, now an assistant professor of microbiology, have not been studied in the context of chronic viral infections.” immunology and molecular genetics at the University of California, The team found six structurally unrelated surface proteins that Los Angeles. played inhibitory roles in CTLs (see Table 1, “Targeting inhibitory Although no patents were filed on the discoveries in the PNAS receptors to boost immunity”). All had previously been implicated in paper, Brooks noted that he and Oldstone previously sought patents on negative regulation of hematopoietic cells in other immunological set- targeting IL-10 to treat viral infection. Neither Scripps nor Scheringtings such as T cell development. In LCMV-infected mice, the recep- Plough Corp. would disclose the licensing status of the antibodies tors were upregulated in exhausted CTLs but not in naïve T cells. used in the study, although the pharma did provide an antibody used In vitro cell-killing, apoptosis and cytokine-production assays sug- to target IL-10 activity. gested that no one receptor was entirely responsible for CTL inhibition. For example, treating exhausted CTLs with antibodies that blocked Monkey bars virus activation of two receptors, CD160 and CD244, increased overall CTL Going forward, Brooks said he plans to test the combination therapy responses compared with untreated controls or in a more advanced animal model, such as priwith antibodies targeting several of the pathmates infected with simian immunodeficiency “Almost all of the new ways individually. virus (SIV) or HCV. targets we identified have The most promising candidate for an inhibiA study published in Nature has already tested not been studied in the tory role in vivo was lymphocyte-activation the effects of blocking PD-1 in SIV-infected context of chronic viral gene 3 (LAG3; CD223). LCMV-infected mice macaques. That report suggests PD-1 blockade infections.” treated with antibodies against LAG3 had alone could decrease viral load and increase sur—E. John Wherry, higher CTL activity and lower viral load than vival, and potentially could be combined with The Wistar Institute mock-treated controls. conventional antivirals to control HIV. Mice treated with a LAG3 antibody and a The study was led by Rama Rao Amara, PD-1 inhibitor had greater CTL activity and lower viral load than assistant professor of microbiology and immunology at the Yerkes those seen using either treatment alone. National Primate Research Center and the Emory Vaccine Center Wherry noted that it’s unclear how the relative importance of the at Emory University. six potential new targets will shake out in humans. To tackle this quesAmara told SciBX that testing the role of PD-1 and other inhibitory tion, his group’s next step is to characterize the expression pattern of receptors in SIV is likely to be more relevant to therapeutic developinhibitory receptors in patients with HIV or HCV. ment than mouse studies. The discoveries reported in Nature Immunology are not patented. To combat SIV, he noted, “you need all three arms of the acquired immune system,” including CTLs, antibody-producing B cells and Take IL-10 helper T cells (CD4+ cells). Helper T cells ordinarily enhance the Meanwhile, a study in PNAS from researchers at The Scripps Research activity of CTLs and B cells, but in SIV and HIV the helper cells Institute will add one more target to Wherry’s list. The paper details themselves fall victim to the virus, thus keeping the entire immune how IL-10, an immunosuppressive cytokine, works in parallel with system at bay. PD-1 to keep CTLs at bay in viral infection. PD-1 is expressed in all three cell types, and Amara᾽s team found Earlier studies identified high levels of IL-10 as a factor in the estab- that blocking PD-1 with antibodies improved the performance of the lishment of persistent viral infections.4 Little is known about how IL-10 cell types in early and late stages of SIV. and its receptor, IL-10R, accomplish their immunosuppressive effects, Treating SIV-infected monkeys with PD-1 antibodies increased the but the new study shows IL-10 and PD-1 work independently and number of SIV-reactive CTLs in the blood and gut compared with could thus be targeted in combination. monkeys given control antibodies. The reawakening of CTLs led to The authors found that LCMV-infected wild-type mice and IL-10 increases in virus-fighting cytokines such as interferon-γ (IFN-γ) and knockouts had equally high levels of PD-1 and its ligand, CD274 (PD- tumor necrosis factor-α (TNF-α). L1). This suggests that IL-10 is not responsible for the upregulation of Moreover, the PD-1 antibody increased the production of SIV-spePD-1 and PD-L1 that accompanies immune exhaustion and implies cific antibodies compared with that seen in mock-treated controls. that IL-10 and PD-L1 accomplish the same goal through different Amara said the inhibitory effect of PD-1 on B cells had not previously pathways. been appreciated. Indeed, blocking both IL-10 and PD-1 pathways yielded a syn“We show that the B cells express PD-1, and if you block PD-1 these ergistic effect. LCMV-infected mice treated with antibodies against cells proliferate and produce more antibodies,” he said. PD-L1 had viral titers lowered 19-fold compared with those from Rejuvenation of CTL and antibody activity against SIV eventually mock-treated controls, whereas interleukin-10 receptor-α (IL-10Rα; led to lower viral loads. SIV titers initially spiked upward in monkeys IL-10R) antibodies decreased viral titers by 21-fold. Treatment with treated with PD-1 antibodies—presumably due to a burst of cell diviboth antibodies lowered viral titers 130-fold, leaving many mice with sion by disinhibited but SIV-infected helper T cells. However, the spike no detectable virus. tailed off as the CTL and B cell responses kicked in. The study was led by Michael Oldstone, professor of immunology at PD-1 inhibition had a pronounced effect on survival. All nine PD-
SciBX: Science–Business eXchange
Copyright © 2008 Nature Publishing Group
targets & mechanisMS Box 1. Positive inhibition. Although a trio of papers shows the merits of blocking inhibitory receptors to treat viral infections, a paper in the Proceedings of the National Academy of Sciences highlights the potential pitfalls of targeting such receptors on all T cells.7 The study, from a team at University College London (UCL), implicates low levels of cytotoxic T lymphocyte– associated protein 4 (CTLA4; CD152) on regulatory T cells as a factor in rheumatoid arthritis (RA). The findings raise the possibility that therapeutic lowering of CTLA4 signaling might also undermine the Treg response that keeps RA in check. Tregs ordinarily fine tune the behavior of other immune cells. In RA and other autoimmune diseases, the immune system fails to pick up on negative cues from Tregs and goes into overdrive against self-antigens. Whereas CTLA4 activity inhibits most immune cells, the new study suggests that in Tregs, CTLA4 has the opposite effect and actually promotes normal activity. Previous efforts to target CTLA4 have aimed to decrease the receptor’s activity on helper T cells. For example, Orencia abatacept, marketed by
Bristol-Myers Squibb Co. to treat RA, is a soluble fusion of CTLA4 and IgG that sponges up CTLA4’s ligand and thus decreases CTLA4 activity. The UCL team isolated Tregs from RA patients and found lower levels of CTLA4 protein and CTLA4-related signaling activity compared with what was seen in healthy controls. As a result, Tregs from RA patients could not inhibit helper T cells as effectively as could healthy Tregs. In cultured Tregs from RA patients, the team found that normal Treg activity could be restored by adding phorbol 12-myristate 13-acetate (PMA), a laboratory compound that short-circuits the inhibitory signaling pathways that keep CTLA4 levels low in RA. Although it is still unknown whether Treg activity is affected by indirect antagonists such as Orencia, UCL’s Michael Ehrenstein told SciBX that the findings bode poorly for therapies that directly inhibit CTLA4. Ehrenstein is professor of experimental rheumatology and a senior author on the PNAS paper. Ehrenstein noted that an earlier knockout study showed that mice lacking CTLA4 in Tregs are susceptible to autoimmune disease.8
“Simply blocking CTLA4 with an antibody would make things worse” for autoimmunity, he added. “This research puts a note of caution to the large scale immunosuppressive therapy that is given to RA patients,” agreed Claudia Mauri, senior lecturer in medicine at UCL and also a senior author on the paper. “Signaling through CTLA4 actually upregulates immunosuppressive mechanisms.” Although PMA is unsuitable as a therapeutic agent because of its broadly disruptive effects on cellular signaling, Mauri believes that other ways of inducing CTLA4 in Tregs could be a reasonable therapeutic strategy. “We want to devise a lentivirus that upregulates or sustains the expression of CTLA4 on Tregs,” said Mauri. Therapy of this kind would most likely involve taking Tregs out of the patient, transfecting them with extra CTLA4 and putting them back into circulation, she added. Mauri and Ehrenstein’s findings have not been patented, but the team has recently reported and sought patents on an in vitro method to stratify RA patients based on their Treg responsiveness.9 —LZO
1 antibody–treated monkeys were alive and healthy at 125 days post- striking, viral titers eventually crept back up, perhaps because rapidly treatment, whereas four of five monkeys treated with a control antibody mutating virus can escape T cell–mediated and antibody-mediated died. immunity. Wherry said Amara᾽s study is “very important for the field since it Amara agreed that mutational resistance by the virus may mean that both confirms and extends the PD-1 blockade approach to primates and antibodies raised by disinhibited B cells are only temporarily effective. SIV. This is a very significant advance, both in terms of the observation He said the problem potentially could be avoided by raising anti-SIV that the treatment was well tolerated by the monantibodies at earlier stages of infection or by keys and the improvement in immunity.” combining PD-1 therapy with standard small “We show that the B cells Amara told SciBX the data in the Nature paper molecule antivirals, which could quash the virus express PD-1, and if you bode well for the development of PD-1-directed before it gets a chance to mutate. block PD-1 these cells adjuvants for conventional antivirals, which The work in the Nature study is the subject of proliferate and produce could be useful in treating HCV as well as HIV. a pending patent, but is not available for licensmore antibodies.” He plans to develop a humanized version of the ing, according to an Emory spokesperson. —Rama Rao Amara, PD-1 antibody and to test the effects of blocking “The SIV paper is very significant,” said Emory University PD-L1. Nils Lonberg, SVP and scientific director at “In theory, it’s a viable approach to chronic Medarex. “For me, the most dramatic thing is viral infection. These data support further exploration” in larger monkey the last figure, where they show the reduction in viral load and increase trials, said Mingjun Huang, executive director of antiviral drug discovery in survival.” at Achillion Pharmaceuticals Inc. Medarex has at least four mAbs in clinical and preclinical developAchillion is developing small molecule therapeutics to treat HCV and ment that target inhibitory receptors described in the three studies. The HIV. The company’s lead compound is elvucitabine, a nucleoside reverse company’s lead product, in development with Bristol-Myers Squibb Co., transcriptase inhibitor that has completed Phase II trials to treat HIV. is ipilimumab (MDX-010), a CTLA4 antagonist in Phase III trials for Huang said that although the effect of PD-1 blockade on survival was melanoma and Phase I and II trials for other cancer indications. SciBX: Science–Business eXchange
Copyright © 2008 Nature Publishing Group
targets & mechanisMS According to Lonberg, cancer and chronic an associate professor of immunology at Scripps “The most dramatic thing viral infection are “analogous” as far as CTL who was not involved in the studies. “These is the last figure, where response is concerned—both tumor cells and braking systems are there for a reason,” he said. they show the reduction in virus-infected host cells thwart the immune There is also debate about which viral indiviral load and increase in system, and reinvigorating CTLs with inhibication would respond best to the disinhibition survival.” tory receptor blockade could thus be effective strategy. Wherry and Brooks think HCV would —Nils Lonberg, Medarex Inc. in both cases. As an example, he noted that be ideal because of its restriction to the liver and Medarex’s MDX-1106 PD-1 mAb, partnered its lack of a latent reservoir. In contrast, hidden with Ono Pharmaceuticals Co. Ltd., is in Phase I trials for both cancer pools of HIV in a variety of cell types have stymied efforts to completely and HCV. eliminate the virus.6 Medarex also is exploring other inhibitory receptor targets. In August, However, Achillion’s Huang believes that HCV would be prone to the company started an open-label Phase I trial of MDX-1105, a mAb the same mutational escape mechanisms as HIV, and that HBV, which that targets PD-L1, to treat cancer. The company also has a preclinical is more genetically stable, might be a better candidate for elimination by program targeting LAG3 to stimulate CTL activity. newly vigilant CTLs. Regardless of indication, Lonberg acknow-ledged that the best use of Combination therapy inhibitory receptor blockers would likely be in combination with small Targeting the new receptors in combination with PD-1 or PD-L1, or molecule antivirals. “Antiviral drugs can tamp down replication, but it’s together with small molecule antivirals, could bypass the concerns about the immune system that sweeps the floor for the last of the virus. So autoimmunity that have dogged CTLA4. I think there’s a real opportunity for immunomodulatory therapies to Wherry said that blocking some of the other receptors identified by improve the efficacy of conventional antiviral therapies.” his group and the Scripps team could bypass the autoimmune side effects Osherovich, L. SciBX 1(45); doi:10.1038/scibx.2008.1092 of CTLA4 blockade that may make it an unsuitable target in chronic viral Published online Dec. 18, 2008 infections. REFERENCES Low levels of CTLA4 have been implicated in the rampant autoimmu- 1. Blackburn, S. et al. Nat. Immunol.; published online Nov. 30, 2008; doi:10.1038/ni.1679 nity that leads to rheumatoid arthritis (see Box 1, “Positive inhibition”). Contact: E. John Wherry, The Wistar Institute, Philadelphia, Pa. Side effects of CTLA4 antagonists in trials for cancer include autoime-mail:
[email protected] mune disorders such as irritable bowel disorder (IBD) and a pigment 2. Brooks, D. et al. Proc. Natl. Acad. Sci. USA; published online Dec. 8, 2008; doi:10.1073/pnas.0811139106 disorder called vitiligo. Contact: Michael B.A. Oldstone, The Scripps Research Institute, In April, Pfizer Inc. discontinued a Phase III trial of its tremelimumab San Diego, Calif. (CP-675206) CTLA4 antagonist in advanced melanoma, citing autoime-mail:
[email protected] Contact: David G. Brooks, University of California, Los Angeles, Calif. mune side effects and nonsuperiority to standard chemotherapy.5 The e-mail:
[email protected] pharma company is continuing Phase II trials in other cancer indica3. Velu, V. et al. Nature; published online Dec. 10, 2008; tions. doi:10.1038/nature07662 Contact: Rama Rao Amara, Emory University, Atlanta, Ga. Lonberg said Medarex’s MDX-010 has elicited inflammatory reace-mail:
[email protected] tions in certain cancer patients, but that these adverse events could be 4. Brooks, D.G. et al. Nat. Med. 12, 1301–1309 (2006) controlled by careful monitoring and prompt treatment with steroids and 5. Ribas, A. et al. Oncologist 12, 873–883 (2007) Remicade infliximab. Remicade, a mAb that blocks TNF-α, is marketed 6. Alexaki, A. et al. Curr. HIV Res. 6, 388–400 (2008) 7. Flores-Borja, F. et al. Proc. Natl. Acad. Sci. USA; published online by Johnson & Johnson to treat autoimmune disorders. Nov. 26, 2008; doi:10.1073/pnas.0806855105 “We have a very carefully calibrated protocol for dealing with these Contact: Michael R. Ehrenstein, University College London, London, U.K. events, so we’re very comfortable with the safety of the drug,” he said. e-mail:
[email protected] Brooks said his team did not see problems with autoimmunity in their Contact: Claudia Mauri, same affiliation as above mouse studies. Similarly, Amara said that in his monkey study, autoime-mail:
[email protected] 8. Wing, K. et al. Science 322, 271–275 (2008) munity did not appear to be a problem. However, Brooks did say such side effects are a concern for future 9. Nadkarni, S. et al. J. Exp. Med. 204, 33–39 (2007) clinical trials, especially ones involving multiple courses of therapy. COMPANIES AND INSTITUTIONS MENTIONED “When you get out in a bigger population where treatments need to be Achillion Pharmaceuticals Inc. (NASDAQ:ACHN), New Haven, Conn. Bristol-Myers Squibb Co. (NYSE:BMY), Princeton, N.J. repeated, there might be problems with autoimmunity,” he said. Emory University, Atlanta, Ga. Brooks and Wherry both said further animal studies are needed to Johnson & Johnson (NYSE:JNJ), New Brunswick, N.J. work out which combination of targets will strike the balance between Medarex Inc. (NASDAQ:MEDX), Princeton, N.J. antiviral response and autoimmunity. According to Wherry, the most Ono Pharmaceuticals Co. Ltd. (Tokyo:4528; Osaka:4528), Osaka, Japan Pfizer Inc. (NYSE:PFE), New York, N.Y. suitable target may vary from person to person, depending on genetic Schering-Plough Corp. (NYSE:SGP), Kenilworth, N.J. makeup and the state of the immune system, and may differ for systemic The Scripps Research Institute, La Jolla, Calif. University of California, Los Angeles, Calif. infections like HIV compared with localized ones like HCV. University College London, London, U.K. The normal role of the new receptors in slowing down CTL response The Wistar Institute, Philadelphia, Pa. against self-antigens needs to be explored further, said Dorian McGavern, Yerkes National Primate Research Center, Atlanta, Ga. SciBX: Science–Business eXchange
Copyright © 2008 Nature Publishing Group
